BMC Medical Education

BACKGROUND: As artificial intelligence (AI) enters clinical practice, concerns about deskilling and never-skilling make the independent assessment of clinical competence increasingly important. The AI-Standardized Clinical Examination (ASCE) uses AI to simulate and score standardized clinical encounters, potentially overcoming some of the logistical and standardization constraints of conventional objective structured clinical examinations (OSCEs). Whether it provides valid and reliable assessment at scale is unknown. METHODS: In this prospective multicenter cohort study, final-year medical students from six French medical schools completed a 10-station ASCE six weeks before the national OSCE session. GPT-4o simulated standardized clinical partners; transcripts were subsequently scored by GPT-4.1 against faculty-defined rubrics. The primary outcome was the correlation between ASCE and national OSCE scores. Secondary outcomes were inter-rater reliability of expert transcript scoring, alignment of ASCE scores with expert ratings, and internal consistency. RESULTS: Among 948 enrolled students, 588 had paired ASCE and national OSCE scores. ASCE scores were correlated with national OSCE performance (Spearman's {rho}, 0.57; 95% CI, 0.51 to 0.63), with less than 1% of residual variance attributable to center (intraclass correlation coefficient, 0.01). Eight experts scored 20 randomly selected ASCE transcripts; Krippendorff's was 0.77 overall, 0.80 for clinical skills, and 0.70 for communication and attitude skills. Adding GPT-4.1 as an evaluator reduced by 0.02, and its scores did not differ significantly from those of individual experts. Internal consistency was 0.67 across 10 stations and was projected to approach 0.80 with 20 stations. CONCLUSIONS: Scores from an examination delivered and scored by AI were associated with performance on a national high-stakes OSCE, with minimal center-level variation, tentative agreement among expert raters, and automated scores that aligned with expert ratings. These results support AI delivery and scoring of clinical examinations, with expert review of automated scores.

Objectives: Health professions students often receive limited training on how to adjust their communication for patients whose first language is not English (LX), despite the risks for diagnosis, treatment, and patient trust. Based on Communication Accommodation Theory, we developed a brief module to improve students' awareness and knowledge of strategic communication adjustments for LX patients. Methods: We tested feasibility in a single-group, pre-post pilot study guided by the Kirkpatrick framework with undergraduate (n=22) and graduate (n=41) speech-language pathology and psychology students in Massachusetts and North Carolina. The module included instructional content, video and audio examples, interactive activities, and a podcast. Learning was assessed with knowledge checks, and self-reported facilitative and resistant accommodation with the Communication Accommodation Behavior Scale (CABS). After training, students rated their confidence and likelihood of using accommodation strategies. Results: At both sites, facilitative accommodation increased and resistant accommodation decreased significantly, with a larger decrease at the undergraduate site. Knowledge-check accuracy was high at both time points, suggesting a ceiling effect. Students reported moderate- to-high confidence in adapting their speech and high likelihood of using a learned strategy in future clinical encounters. Conclusions: A brief module can be feasibly delivered across institutions and is associated with short-term improvements in self-reported communication accommodation. Practice implications: We recommend future work evaluate whether these changes translate into clinical behavior long-term and refine knowledge-check items to better capture learning gains.

Background: Parenthood in medicine is an experience fraught with challenges that begin in the pre-conception phase and persist as physician-parents navigate their dual role. Methods: This study aims to understand the experiences of physician-parents (both childbearing and non-childbearing) returning to work after parental leave. Physicians, including residents, affiliated with the University of Toronto Department of Medicine internal medicine division who had a child within the last ten years were invited to participate in a self-administered electronic survey from June2022 to 30June 2023. Data were summarized using descriptive statistics. Results: Twenty-nine respondents took part in the survey; 79% were staff physicians while the remaining were trainees at the time of the survey. At the time their first child was born, 48% were staff members and 52% were trainees. The median duration of parental leave was 5 months (range 1-12 months). Among those for whom lactation was applicable, 76% reported that they continued their lactation efforts on returning to work, while 24% chose not to. Thirty five percent (35%, n=6/17) of respondents cited that they were not provided with duty hour modifications when required. Many respondents (46%) felt guilty about burdening their co-workers for requesting accommodations for childcare. When requested to rate their experiences of balancing their personal and professional lives on a scale of 1-10 (1 being impossible to maintain and 10 being extremely easy to maintain), most respondents (66%) chose a rating of 5 or less. Conclusion: This study supports previous findings that physician-parents face a myriad of challenges integrating their roles in medicine and parenthood. A change in culture is required to support and normalize parenthood for physicians and foster sustainability and productivity in academic medicine.

Abstract Background: Large language models (LLMs) are increasingly explored as adjunct tools in graduate medical education, but their reliability for answering discipline-specific certification examinations remains incompletely characterized. In dentistry in China, the national dental examination serves as a milestone for stomatology trainees. Prior studies predominantly rely on single metric, limiting reproducibility and granularity. Objective: To systematically evaluate the performance of five LLMs in answering the national standardized training examination for dental residents. Method: A total of 1,318 questions covering the six dentistry disciplines were selected. Five LLMs (DeepSeek-V3, DeepSeek-R1, Qwen3-32B, Claude 3.5, and GPT-4o) were evaluated via a single-response protocol on the Dify-based Chatflow framework. Performance combined human verification with automated scoring using the Retrieval-Augmented Generation Systems (RAGAS) framework across seven dimensions. Tasks involving question classification, key-point categorization, and instructional requirement summarization were designed to enable a horizontal comparison of models' cognitive capabilities, assessed using inter-model consistency scores. Results: DeepSeek-R1 achieved the highest correct choice rate (80.90%, p < 0.001) and answer correctness score (0.853, p < 0.001), significantly outperforming the others. All models exhibited limitations, including information insufficiency, redundancy, and hallucination-related issues (lowest 44.20% in DeepSeek-R1). In classification and summarization tasks, Claude 3.5 showed higher consistency scores (186/181) than Qwen3-32B (152/151). Conclusion: Evaluated LLMs demonstrated good performance on dental examination questions and in executing tasks. DeepSeek-R1 consistently outperformed the others. While LLMs showed potential as supportive tools in dental education, further refinement and validation are needed before formal adoption in graduate medical educational and clinical training settings.

Background: Artificial intelligence (AI) is increasingly used in healthcare, but data on its real-world use and perceptions among obstetricians and gynecologists remain limited. Objectives: To assess familiarity, use, perceptions, and expectations regarding AI among physicians practicing obstetrics and gynecology in France. Study Design: We conducted a cross-sectional survey between June 2 and July 30, 2025, mainly distributed through professional organizations or associations and social media. Participation was voluntary and self-selected. AI use was defined as any use of an AI tool in routine clinical practice, regardless of type or frequency. Comparisons were made between AI users and non-users and between junior and senior physicians. Results: Among 469 respondents, representing approximately 5% of the target population, 199 of 466 (42.7%) reported AI use in practice; three respondents were excluded from this analysis because of discordant questionnaire responses. AI use was associated with male gender and professional status, with residents more represented among users. Junior physicians reported greater AI use than senior physicians and expressed greater concern regarding medical confidentiality. Current applications primarily involved administrative tasks, medical writing, literature retrieval, and translation; anticipated future uses centered on clinical decision support and medical data interpretation. Perceptions of several AI applications differed between groups. Conclusions: AI was reported to be used by a substantial proportion of respondents, primarily for supportive and informational tasks. These findings support further evaluation of training, validation, governance, and safe implementation of AI in obstetrics and gynecology.

Abstract Background: Cyprus has transitioned into a high-alert zone for mosquito-borne viruses. This necessitates a proactive assessment of its future physicians. Following the 2022 re-establishment of Aedes aegypti in Larnaca and Aedes albopictus in Limassol, Cyprus, the island is now biologically primed for local outbreaks of dengue, Zika, and chikungunya, alongside Culex-borne West Nile virus. The study assessed medical students' knowledge and risk perception. Methods: A cross-sectional online survey administered the validated MosquitoWise Health Belief Model instrument to undergraduate-entry and graduate-entry students at a university in Cyprus in mid-2026. Results: Of 1023 students, 119 responded (11.6%), with pre-clinical students overrepresented (p<0.001). Mean knowledge was 5.43/9, stronger on mosquito biology (72.5%) than virus knowledge (54.2%). Median severity was 6.00/7 against susceptibility 4.25/7. Knowledge differed by study stage (p<0.001), correlated with severity (rs = 0.399) but not susceptibility. Conclusions: The study establishes a critical baseline for pre-clinical medical students in Cyprus. Overall, participants performed better on mosquito biology when compared to mosquito-borne viruses and how they spread. Our findings highlight the value of integrating theoretical microbiology with clinical practice in medical education to enhance physicians' preparedness for the changing landscape of newly emerging infections in Cyprus and across Europe as a whole. Key Words: Arboviruses, Cross-Sectional Studies, Cyprus, Health Belief Model, Medical Students, Risk Perception.

Background While overall application volume by ethnic minority doctors into UK postgraduate medical training has expanded significantly, workforce diversity varies substantially across specialties. Previous reports have suggested that doctors from certain ethnic backgrounds and international medical graduates (IMGs) experience lower likelihoods of securing training posts, but granular longitudinal evaluations across all specialty training levels and recruitment stages remained limited. Therefore, this study evaluates longitudinal trends in self-reported ethnicity and country of qualification across postgraduate medical training in England between 2021 and 2025, maps specialty representation, and tests the hypothesis that recruitment success rates differ significantly between ethnic minority and White baseline applicants. Methods We conducted a retrospective national longitudinal analysis of mandatory NHS specialty recruitment data across England from 2021 to 2025, encompassing over 300,000 applications across ST1/CT1 (entry-level) and ST3+ (higher specialty) pathways. Applicants were categorised into census-aligned broad and granular ethnic groups, alongside IMG classifications. Unadjusted odds ratios (ORs) with 95% confidence intervals were calculated using White graduates and internal sub-ethnic baselines as comparators. Specialty-level ecological correlations (Spearman's rho) were performed to evaluate whether international workforce reliance moderated ethnic recruitment disparities. Results Between 2021 and 2025, applicant volume expanded rapidly, with ethnic minority applicants and IMGs comprising numerical majorities. However, Asian and Black applicants experienced substantially lower odds of receiving training offers compared to White baseline applicants across multiple high-volume specialties (e.g., emergency medicine and anaesthetics at ST1, where Black applicants faced multi-fold reductions in success odds: OR 0.03 to 0.05, p<0.001). Granular analyses revealed notable intra-ethnic heterogeneity, including lower success among Pakistani and Bangladeshi applicants compared to Indian applicants. Furthermore, UK graduates maintained higher offer success than IMGs across most specialties. Ecological correlations demonstrated that higher international applicant reliance was moderately and significantly associated with wider recruitment disparities at higher training levels for Black applicants {rho}=-0.43, p=0.019). Conclusions Centralised medical recruitment in England exhibits some persistent ethnic and international qualification disparities despite rising application diversity. The findings highlight the necessity of disaggregating ethnic groups and addressing subjective evaluation vulnerabilities within multi-stage recruitment processes to foster equity. Trial registration: Not applicable

Background: Neuroanatomy poses a significant challenge for Psychology students due to its spatial and conceptual complexity. Educational approaches that enhance the relevance and visualization of neuroanatomical content may improve students learning experiences. This study implemented a tractography-based activity focused on the virtual dissection of the arcuate fasciculus, a major white matter pathway, in undergraduate Psychology students and examined the relationships between perceived learning and students perceptions of utility, difficulty and handling, and organizational aspects of the activity. Methods: First-year undergraduate Psychology students participated in a two-session tractography-based activity combining instruction on white matter anatomy and diffusion tractography with a hands-on virtual dissection of the arcuate fasciculus using research-grade software routinely employed in neuroscience research. Following the activity, students completed an anonymous questionnaire assessing perceived learning, utility, difficulty and handling, and organizational aspects of the activity. Pearson correlations, multiple regression analyses, and relative importance analyses were performed. Results: Sixty-eight students completed the questionnaire. Students reported generally positive perceptions of the activity across the evaluated dimensions, with perceived learning receiving the highest mean score (M = 3.44, SD = .85). Perceived utility showed the strongest association with perceived learning (r = .64, p < .001). The regression model explained 41% of the variance in perceived learning (R2 = .41, adjusted R2 = .38, p < .001). Perceived utility was the only significant predictor in the model ({beta} = .59, p = .001), accounting for 67.9% of the explained variance. Conclusions: The findings support the feasibility of integrating authentic neuroimaging tools into undergraduate neuroanatomy teaching. Students who perceived the activity as more useful also reported higher perceived learning outcomes, with perceived utility emerging as the strongest predictor of perceived learning. In contrast, perceived difficulty and handling, and organizational aspects did not make significant independent contributions. These results suggest that students perceptions of educational relevance may play an important role in technology-enhanced STEM learning experiences.

Background: System One Model is designed for structured decision-making and can provide probabilistic outputs, but their performance in domain-specific physical therapy tasks has not been established. Objective: To evaluate the performance and potential utility of Jev, a System One Model, for professional knowledge-based selection tasks in physical therapy by comparing it with GPT-4o. Methods: A total of 380 multiple-choice questions from the 2024 and 2025 Korean Physical Therapist Licensing Examinations were independently submitted to Jev and GPT-4o. The outcomes included answer accuracy, subgroup performance, output validity, API response latency, token usage, estimated API cost, and Jev's prediction uncertainty. A retrospective probability-based model-cascading simulation was also performed. Results: Jev correctly answered 282 questions (74.2%), whereas GPT-4o answered 326 (85.8%), a difference of -11.6 percentage points. Jev produced valid structured responses for all questions, whereas GPT-4o produced three invalid responses. Jev's selected-answer probabilities discriminated correct from incorrect answers (AUROC = 0.866), and accuracy reached 99.3% when probabilities were greater than or equal to 0.90. The median API response latency was 0.94 s for Jev and 1.06 s for GPT-4o, with estimated costs of US$0.008 and US$0.161, respectively. At a retrospective cascading threshold of 0.70, accuracy was 86.6% with 175 simulated GPT-4o calls. Conclusion: Jev showed lower overall accuracy than GPT-4o but provided reliable structured outputs, informative prediction probabilities, lower latency, and lower estimated cost. Its probabilistic outputs warrant further prospective evaluation for selective model escalation and decision-support workflows in physical therapy.

Background: Large language models (LLMs) configured as virtual patients may expand structured practice of clinical reasoning in health professions education. Before testing effectiveness, it must be established whether students engage with them and whether the resulting scores are complete, reliable, and related to performance with real patients. Our earlier pilot randomized trial found very low engagement with a less structured LLM-based approach. Objective: This study aimed to assess the feasibility (adherence, acceptability, and data completeness) of a revised, structured LLM-based clinical reasoning laboratory (Laboratorio de Inteligencia Artificial [LIA]) in undergraduate physical therapy education and to explore its measurement properties, including its association with clinical judgment with a real patient assessed with the Lasater Clinical Judgment Rubric (LCJR). Methods: In this single-arm pilot feasibility study, 27 third-year physical therapy students at a Spanish university completed, in 2024-2025, 3 virtual patient cases (lumbar, cervical, and thoracic spine pain) simulated by ChatGPT (GPT-4) over 4 weeks, each twice. ChatGPT scored each attempt from 0 to 4 on the same 10 dimensions of the Lasater Clinical Judgment Rubric (LCJR; 0-40 points) that human assessors applied (1-4 per dimension; 10-40 points) during a real patient encounter. Feasibility outcomes were case completion, data completeness, and satisfaction (0-10). Measurement outcomes were internal consistency and convergent validity (Spearman correlations between model-generated and human LCJR scores). Variability estimates informed the sample size of a future trial. Results: All 27 students were assessed with the LCJR; 22 (81%) completed at least 1 LIA case. Both attempts of the lumbar, cervical, and thoracic cases were completed by 78%, 63%, and 74% of students, respectively. Satisfaction data were available for 15 students (56%; mean 6.0, SD 1.3), and missing secondary data clustered in the third rotation. Internal consistency was excellent for the LCJR (alpha=.96) and ranged from .43 to .90 for model-generated ratings. Model-generated LIA scores clustered near the maximum (mean 36.0, SD 1.4). Among completers, LIA scores correlated weakly with the LCJR (rho=0.25, 95% CI -0.20 to 0.61; rho=0.44, 95% CI 0.05-0.71, when noncompleted attempts were scored as 0). Grade point average correlated almost perfectly with the LCJR (rho=0.96; n=22); after adjustment for it, the LIA-LCJR association disappeared (partial rho=-0.10). A future trial would require approximately 39 students per group (80% power, conservative variance). Conclusions: Delivering the LIA within the curriculum was feasible, but clinical judgment ratings generated by ChatGPT did not reflect those of human assessors with a real patient: they clustered near the maximum and showed no convergent validity once noncompletion was separated from performance. LLM-generated ratings should not be used as outcome or assessment measures without validation against human raters. Case calibration, completion tracking, and clarification of the GPA-LCJR relationship are also needed before a definitive trial. Trial Registration: ClinicalTrials.gov NCT06809634; https://clinicaltrials.gov/study/NCT06809634

Objective To evaluate the information quality, reliability, readability, and content deficits of ChatGPT, DeepSeek, and Doubao in answering frequently asked questions about bee stings. Methods Twenty-five high-frequency patient questions were selected based on Google Trends, domestic health platforms, and allergy guidelines, and submitted to the three LLMs, yielding 75 responses. Two emergency physicians independently assessed quality using GQS, mDISCERN, QAMAI, EQIP, and PEMAT-U/A. Readability was evaluated using six indices (ARI, FRES, GFI, FKGL, CLI, SMOG). Content deficits were classified qualitatively. Twenty emergency department patients evaluated the comprehensibility and credibility of AI responses. Results Significant differences were found among the three models on all quality metrics (all P < 0.01). DeepSeek achieved the highest mDISCERN score (median 21.00); ChatGPT led on QAMAI (25.00) and EQIP (75.00%); Doubao ranked lowest on all quality indicators but highest on PEMAT-U (73.00%). No model met the sixth-grade reading level standard. Eleven content deficits were identified: outdated medical advice (45.45%), omissions (36.36%), and hallucinations (18.18%). Doubao received the highest patient preference (45.0%). Conclusion The three LLMs exhibited distinct strengths: DeepSeek in source reliability, ChatGPT in overall quality, and Doubao in understandability. All models had excessively high readability levels, and outdated medical advice was the predominant content deficit. Objective To evaluate the information quality, reliability, readability, and content deficits of ChatGPT, DeepSeek, and Doubao in answering frequently asked questions about bee stings. Methods Twenty-five high-frequency patient questions were selected based on Google Trends, domestic health platforms, and allergy guidelines, and submitted to the three LLMs, yielding 75 responses. Two emergency physicians independently assessed quality using GQS, mDISCERN, QAMAI, EQIP, and PEMAT-U/A. Readability was evaluated using six indices (ARI, FRES, GFI, FKGL, CLI, SMOG). Content deficits were classified qualitatively. Twenty emergency department patients evaluated the comprehensibility and credibility of AI responses. Results Significant differences were found among the three models on all quality metrics (all P < 0.01). DeepSeek achieved the highest mDISCERN score (median 21.00); ChatGPT led on QAMAI (25.00) and EQIP (75.00%); Doubao ranked lowest on all quality indicators but highest on PEMAT-U (73.00%). No model met the sixth-grade reading level standard. Eleven content deficits were identified: outdated medical advice (45.45%), omissions (36.36%), and hallucinations (18.18%). Doubao received the highest patient preference (45.0%). Conclusion The three LLMs exhibited distinct strengths: DeepSeek in source reliability, ChatGPT in overall quality, and Doubao in understandability. All models had excessively high readability levels, and outdated medical advice was the predominant content deficit.

Background: The rapidly evolving United States (U.S.) health and science policy landscape in 2025 may have heterogeneous effects on academic medicine, biomedical research, physician workforce development and patient care. Research Question: How do academic pulmonary and critical care medicine (PCCM) physicians perceive U.S. health and science policy changes and which changes do they consider most influential on research, trainee education and clinical care? Study Design and Methods: We conducted a prospective survey from March 27, 2026 through April 29, 2026 among selected faculty at U.S. academic PCCM programs as part of the Alliance of Scholars and Professionals for Innovation, Research, and Education in Pulmonary and Critical Care Medicine (ASPIRE-PCCM). Results: Ninety-four of 99 ASPIRE-PCCM panel members completed the survey (response rate 95%), representing 18 of 19 ASPIRE-PCCM wave 1 institutions. All respondents reported the rate of changes in U.S. federal health and science policy in the first half of 2025 as 6 or higher on a scale of 1 ("typical pace of change") to 10 ("most change in my lifetime"), with 89 (96%) evaluating the changes as harmful or very harmful. Respondents identified cancellation of research grants, disruptions and changes to NIH scientific review processes, and changes to visa and immigration policies as key factors affecting early career recruitment. Reported negative effects on quality of care included increased vaccine hesitancy, changing reimbursement structures, and diminished research capacity. State-level and non-governmental efforts, including professional society advocacy, institutional bridge funding and personal advocacy efforts were identified as partially mitigating harms. Despite these concerns, 72% of respondents reported that would still pursue careers in PCCM. Interpretation: Academic pulmonary and critical care physicians reported challenges following recent U.S. health and science policy changes that may be mitigated, but not neutralized, by institutional, professional-society, and individual actions.

Background Interruptions are common in hospital outpatient departments (OPDs) and are linked to error. Many are clinically necessary, so the aim is to regulate them. Most interventions target the individual clinician. Little is known about how doctors manage interruptions in practice or what makes their strategies work. Methods Qualitative study using focus groups and interviews with 42 doctors working in UK National Health Service OPDs (19 sessions, September to November 2025). Participants were purposively sampled by clinical experience and self-reported patient complexity. Data were analysed using Framework Analysis, informed by Cognitive Load Theory and Social Identity Theory. Results We developed four themes. (1) Pre-emptive protection of attention. Doctors withdrew, signalled unavailability and set thresholds for interrupting. These depended on colleagues learning and sharing the norm, which staff rotation eroded. (2) Sharing the burden through team coordination. Designated roles and time-limited rotas moved interruptions to colleagues with capacity set aside for them and made interrupting legitimate. (3) Real-time cognitive repair. Doctors asked interrupters to wait while they secured the task, and rebuilt the consultation on return, sometimes with the patient's help. Asking a colleague to wait depended on seniority or team membership; some trainees described difficulty doing this with senior colleagues. (4) Organisational recognition as the precondition for sustainable regulation. Where supervision was built into job plans, the interruptions it generated were expected and doctors held capacity in reserve for them. Advice, vetting and telephone queries were rarely counted as work. Conclusions Each strategy served a cognitive purpose, and social and organisational conditions decided whether it was perceived to work. Interventions could make norms shared and roles explicit, count interruption work in job plans and keep routes open for junior staff to raise concerns. Evaluations should ask who is able to use them.

Background: Disaster medicine has been a formal curricular recommendation in German health professions education since 2006, and comparable recommendations exist internationally. Evidence on implementation is confined almost entirely to human medicine, rests on faculty self-reports or small student samples, and measures delivery rather than retention. Whether curricular exposure is associated with willingness to serve has not been adequately examined in interprofessional samples. Objective: We aimed to characterize disaster medicine education across 7 health professions in Germany, distinguishing delivery from retention, to identify profession-specific gaps and training priorities, and to examine how educational and experiential factors relate to willingness to serve. Methods: We conducted an anonymous nationwide online cross-sectional survey of students and trainees across 7 health professions in Germany between July 9 and August 25, 2026, recruited through institutional distribution lists. The interprofessionally developed instrument comprised 48 items. We assessed delivery and retention of 12 disaster medicine topics (4-point scale), self-rated confidence in 22 clinical and nursing skills (6-point scale), desired content, formats, motivators and barriers, and formal and spontaneous willingness. For all rating scales, higher values denote the more favorable level. We derived a delivery-retention gap and a want-deficit index, examined the dimensional structure of the competence battery, tested for differential item functioning by prior civil protection engagement, and modeled willingness with a cumulative link model. Results: Of 8133 respondents (completion rate 87.5%), 33.1% had never been taught disaster medicine and 25.0% only in passing; 22.1% had comprehensive teaching including practical exercises. Across the 12 topics, the proportion able to retrieve content confidently never exceeded 23.9%; the delivery-retention gap reached 47.3 percentage points. More than half (54.6%) had never taken part in a practical exercise. Competence formed more than 1 dimension: 18 acute-technical items loaded 0.54-0.86 on the first factor, 4 relational and nursing items 0.14-0.28. Prior civil protection engagement was associated with higher confidence on every acute-technical item (Cliff delta 0.26-0.55) and lower confidence on all 4 relational items (-0.22 to -0.09). Willingness was high and homogeneous across groups (78.7% formal, 84.1% spontaneous), whereas competence varied by about 2 scale points. Adjusted, active civil protection engagement (odds ratio 2.49, 95% CI 2.12-2.94) and practical exercise experience (odds ratio 1.78, 95% CI 1.54-2.05) were the strongest predictors; curricular exposure contributed weakly (odds ratio 1.06 per SD, 95% CI 0.99-1.14). Conclusions: Two decades after the curricular recommendation, disaster medicine is largely absent from health professions education in Germany, and what is delivered is not retained. Willingness is nevertheless high and largely decoupled from competence, so qualification and recruitment require separate instruments. Curricula should be prioritized by the coincidence of demand and deficit rather than by coverage, built for retention, and should treat relational and nursing competence as a distinct learning objective.

Background AI use is growing by HPE scholars, and disclosure of that use is required by scientific ethics and journal policy. In spite of growing unease about the meaning, and meaningfulness, of AI-use disclosures in published manuscripts, little is known about how authors represent their AI use in published disclosure statements. Without such understanding, we are ill-equipped to redirect disclosure practices for increased sufficiency and sincerity, undermining efforts towards transparency in the era of AI-supported scholarship Methods In this descriptive qualitative study, 18 HPE researchers with experience disclosing generative AI use in publications from 2025-2026 were interviewed using Zoom. Data collection and thematic analysis proceeded iteratively, exploring participant experiences with AI use, their disclosure decisions, and evolving disclosure practices. Results Participants reported a wide range of experience writing AI-use disclosures (from 1-16 disclosures/participant); in total 31 published disclosures were shared by participants and discussed during the interviews. Our analysis identified four themes: A spectrum of AI uses; Multi-factorial disclosure decisions; Discrepancies between actual and disclosed use; and Emotions around disclosure. Discussion AI-disclosure takes place at the intersection of shared ethical principles and situated pragmatic concerns. Disclosure is both an intellectual and an emotional practice, changing over time as AI use evolves and journal expectations develop. More generative uses may provoke divisive responses and are related to uncertainty about how to disclose, concern about disclosure penalties, and periodic use-disclosure discrepancies. Understanding how researchers are navigating these complexities may help the HPE community articulate clearer guidance for transparent and meaningful disclosure of AI-use in scholarly work. Reflective methodologies can help bring these issues to the surface of ongoing conversations about AI use.

Objectives To examine the prevalence of interruptions during UK secondary care outpatient consultations, their association with clinician mental workload, and the strategies clinicians use to manage them. Design Nationwide cross-sectional online survey. Setting Secondary care outpatient clinics, United Kingdom. Participants A sample of 465 UK secondary care doctors across a range of grades and specialties, recruited through a medical defence organisation's network. Primary and secondary outcome measures Mental workload (NASA Task Load Index, Raw-TLX); interruption frequency; perceived cognitive impact of interruptions; task resumption difficulty; perceived social effects on colleague interactions and team cohesion; and interruption-management strategy use. Results 61.1% of clinicians reported being interrupted often or very often, and mean overall workload was high (Raw-TLX 66.2, SD 11.5). Interruption frequency, patient complexity and perceived cognitive impact were each associated with higher workload across multiple NASA-TLX subscales. In fully adjusted models, interruption frequency retained independent associations with overall workload (beta = 1.79, p = 0.004) and temporal demand (beta = 0.94, p < 0.001). Task resumption difficulty was associated with worse perceived performance (beta = -0.59, p = 0.014), while perceiving interruptions as socially positive was associated with better performance (beta = 0.75, p = 0.003) and lower frustration (beta = -1.16, p < 0.001). Years of experience showed only small effects, and self-reported strategy use (38.7%) was not consistently associated with lower workload. Exploratory natural language processing of open-text descriptions identified three strategy clusters; cognitive offloading, anticipating and preparing, and blunt responses, with distinct emotional profiles. Conclusions Interruptions were commonly reported during UK secondary care outpatient consultations and were associated with measurable cognitive and temporal burden. The perceived social framing of interruptions was independently associated with frustration and self-rated performance. Binary strategy use showed no consistent workload association, indicating a need for system-level approaches addressing interruption frequency, task resumption support and the social context of interruptions rather than reliance on individual coping.

Background: Palliative care is suboptimal in pulmonary fibrosis. Symptom burden is high despite evidence-based interventions, guidelines, and information resources. The application of existing knowledge and methods to optimally manage palliative and end of life care for people with pulmonary fibrosis in the UK has not been realised. Aim: To establish patient views and understanding of how palliative care services and approaches can better meet the needs of people with pulmonary fibrosis. To identify the current barriers and facilitators to receiving palliative care services. Design: Qualitative exploration of patient views via semi-structured interviews and focus groups. Content analysis was conducted. Setting/Participants: The study was conducted from March to August 2024 within an Interstitial Lung Disease Specialist Centre, based in an NHS acute Trust. Results: Two focus groups and four interviews were conducted, involving nine patients. Four overarching themes were identified as to how palliative care can better meet the needs of this group: (1) establishing mechanisms for identifying and addressing palliative care needs; (2) open channels of communication with healthcare professionals; (3) the importance of a palliative care approach within respiratory, community and specialist services; and (4) support for unpaid/family carers. Conclusions: Patients highlighted the importance of person-centred principles and services that adapted to their needs (such as real-time telephone support), rather than services that imposed their own requirements on the patients. Understanding how existing palliative care provision can meet this standard could help improve the palliative and end of life care for people with pulmonary fibrosis.

Introduction Bubble Continuous Positive Airway Pressure (bCPAP) is an effective intervention for managing respiratory distress syndrome among preterm newborns. However, its successful utilization depends on healthcare providers possessing adequate knowledge and practical skills. In Tanzania, competency-based equipment user training has been introduced to strengthen providers' capacity to use neonatal technologies, including bCPAP. Despite these efforts, there is limited qualitative evidence on how this training influences healthcare providers' competence and routine clinical practice. This study explored the role of competency-based equipment user training in enhancing healthcare providers' skills in utilizing bCPAP devices for newborn care in Dar es Salaam, Tanzania. Methods A phenomenological qualitative study was conducted in three regional referral hospitals implementing the NEST360 programme. Purposive sampling was used to recruit doctors and nurses who had completed competency-based equipment user training. Data were collected through in-depth interviews using a semi-structured interview guide and analysed using Braun and Clarke's thematic analysis approach. Findings Six major themes emerged. Participants described competency-based training as a great approach for improving practical skills, confidence, and clinical decision-making regarding patient selection, device operation, monitoring, and weaning. Participants also reported improved equipment maintenance practices, reduced preventable equipment damage, and stronger collaboration with biomedical engineering professional . The training promoted positive changes in provider behaviour, including greater autonomy, internal mentorship of newly recruited staff, and increased accountability for equipment management. Participants further perceived improvements in neonatal care through earlier initiation of bCPAP, fewer unnecessary referrals, and better management of preterm newborns within their facilities. Conclusion Competency-based equipment user training strengthened healthcare providers' competence and confidence in utilizing bCPAP for newborn care while improving equipment management and interdisciplinary collaboration. Integrating competency-based training into routine newborn care programmes may enhance the effective implementation and sustainability of life-saving neonatal technologies in resource-limited settings.

Gamete donation may widen access to assisted reproductive technology, but donor gametes remain difficult to obtain in Ghana. This institution-based cross-sectional survey assessed knowledge, attitudes and willingness to donate among 380 undergraduate students at Ho Technical University. The sample was allocated proportionately across 24 strata defined by six faculties and four academic levels; within each stratum, students who were present during scheduled classes were recruited consecutively. Data were collected using a self-administered questionnaire and analysed using descriptive statistics, exact tests and exploratory logistic regression. Although 88.2% of participants had heard of gamete donation and 72.4% supported it, only 42.1% said they would donate if approached. Among the 220 participants who were unwilling, 210 (95.5%) selected the response that donation felt wrong to them. Participants who perceived societal attitudes as positive had higher odds of reported willingness than those who perceived them as negative (adjusted odds ratio [aOR] 3.08, 95% confidence interval [CI] 1.34-7.08). Belief that donation could exhaust ones gametes was associated with lower odds of willingness (aOR 0.27, 95% CI 0.15-0.48). The conditional statement that payment would increase willingness was also associated with direct willingness (aOR 7.16, 95% CI 4.25-12.06), but the two measures overlapped conceptually. The findings identify fertility misconceptions, concerns about donation and differing views on compensation that should be considered when developing donor information, counselling and protection measures.

Introduction: Nursing students experience stress arising from demanding academic requirements and emotionally challenging clinical placements, which may affect wellbeing, learning, and programme retention. This study explored perceptions of stress, its impacts, and coping experiences within undergraduate nursing education among nursing students and educators in New Zealand. Methods: A qualitative descriptive study was conducted with six undergraduate nursing students and six nurse educators at a New Zealand university. Data were collected through semi-structured interviews and analysed using inductive reflexive thematic analysis, with interpretation informed by the Transactional Model of Stress and Coping. Results: Three themes were identified: stress from multiple sources, stress shaping wellbeing and learning, and navigating stress through appraisal and coping. Stress arose from academic, clinical, personal, and social demands and affected academic engagement, health, and relationships. Participants described varied responses to stress, including different appraisals of challenging situations and coping strategies such as peer support, self-care, and use of institutional resources. Conclusion: Stress in nursing education reflects a dynamic interaction between individual appraisal, coping resources, and educational and interpersonal contexts. Although findings are limited to one institution, they highlight the importance of accessible, flexible, and culturally responsive support and of addressing modifiable educational stressors alongside strengthening students' coping resources.